Research On Key Issues Of Distributed Virtual Environments

With the development and popularization of computer technology, network technology and virtual reality technology, DVE (Distributed Virtual Environment) system had great progress both in theoretical research and particular application. In order to realize spatial and temporal sharing, different technologies and methods (e.g. client-server setting, DR algorithm, grouping method and special transmission protocol etc.) came into being to increase system's consistency and scalability. However, with the increase of system participant and task interaction, these technologies and methods could not meet DVE system's request.Two challenges existed in spatial and temporal sharing: how to increase system's message filtering capability and how to enhance system's interaction ability. In order to enhance system's interaction ability, spatial and temporal sharing technology should be improved in following areas: increase participant behavior's accuracy and consistency; support multiple participants to operate one entity in same time; be adaptable to different tasks, offer different level of accuracy and consistency according to different types of task; be adaptable to different bandwidth, transfer most important messages according to current bandwidth.The goal of this paper was to present a spatial and temporal sharing framework (including consistency maintenance, grouping method, DR algorithm and special transmission protocol) for DVE system to increase scalability and offer higher interaction ability. Specifically, this paper made following contributions:Time warp based state repairing mechanism. Because traditional DR algorithm had many short time inconsistency and local lag technology could not support multiple participants to operate one entity in same time, time warp based state repairing mechanism was presented. Cooperating with local lag technology, this mechanism could eliminate short time inconsistency and support multiple participants to operate one entity in same time. This mechanism could be used in DVE system which have high request on accuracy and consistency.Fuzzy grouping method. Traditional grouping method could not use all attributes of entity and fuzzy correlation space based approach needed more match computation. Fuzzy grouping method added two new techniques based on the fuzzy correlation space approach: fuzzy consistent relation based weight distribution and grid based preprocessing. This method not only kept high and stable data filtering rates and effective data receiving rates in different situation but although greatly reduced matching number.Knowledge based DR algorithm. Traditional DR algorithm used a fixed threshold. But knowledge based DR algorithm could adjust the threshold according to the task type. This algorithm could not only meet task's request onaccuracy and consistency but although filter most messages and increase scalability.Interactive stream transmission protocol. Interactive stream transmission protocol was presented to meet DVE system's request on scalability and accuracy. This protocol realized selective filtering, transmission latency and jitter control, adaptive adjustment, and not only meet system's request on accuracy and consistency but although filter most messages and increase scalability.Research on network characteristics' effect on task performance. This paper researched the effect of network transmission latency and packet loss on a special task in a DVE system, and got the maximum acceptable transmission latency and packet loss rate in doing this task. The results of this research could be used to define DVE system's network QoS, which should be regarded as a standard in DVE system designing and implementing.Besides, this paper presented two experimental systems which adopt new methods and technologies presented in this paper. One system was JBCVE (Java Based Collaborative Virtual Environment), which was developed to research on spatial and temporal sharing technology used in DVE system having strict request on interaction and...